Big-Data Insights

With the help of WealthEngine, JDRF has cleaned up its big-data business processes and enhanced its fundraising. Here, Megan Martin, former director of data analysis at JDRF (now at American Cancer Society), and Sally Boucher, director of research at WealthEngine, share some data insights.

Know where to start
“The whole idea of big data sounds very intimidating. That in itself presents a stumbling block for a lot of smaller nonprofits,” Boucher says. “You really need to have a system for collecting and storing and accessing data. You cannot really know what you have or need or how you’re going to get there without having a place where you can audit your current data.”

Her recommendation is to work with a consultant or partner to get started and have a central repository to house your data.

Data quality is key
“A lot of nonprofits, especially small nonprofits, are still getting on board with getting and capturing the data they need, so the data is not always as clean as they hope,” Martin says. “All the business processes haven’t caught up yet, so data quality is sometimes a stumbling block.”

WealthEngine did a major survey on data collection and found that the biggest issues are data cleanliness, consistency and accuracy.

Have systems in place
“Systems are absolutely essential for all business processes,” Boucher says. “To execute a strategy, having a system allows you to test different methods, inputs, track progress, evaluate the effectiveness, etc. Only then can you go on and scale what you’re doing.

“Tracking and evaluating are essential. When talking about big data, there are so many moving pieces, so many skill sets required in working with data — from developing infrastructures to handle the data, to people who can extract that into reporting mechanisms to interpret for business needs,” she adds. “You have to have a data strategy and a process for dealing with it. It can’t be done piecemeal.”

Listen to your data
“How does data inform programs? How do you segment based on what the data is telling you? How do you message people differently based on data? Test different approaches, and let the data tell you what works and what doesn’t work,” Boucher says.

Let fundraisers fundraise
“Fundraisers aren’t always data analysts; they’re fundraisers,” Martin says. “When you have limited resources, you’ve got to have people out there doing their jobs.”

So, Martin says, you have to find a specific skill set for data, and getting outside help may be necessary.

“Collecting data, putting pretty charts up — it’s no good unless somebody can interpret it and use it,” she adds. “While the self-service model is great, not everybody can interpret it. It can be overwhelming and confusing to people who don’t understand it.”

Behavioral data is very important
All nonprofits should collect information like giving history, event attendance, interactions with the organization — and you should be tracking it as well, Boucher says.

In addition, look at differentiators such as wealth data, direct-mail or email ask amounts, demographic data, life stage, children vs. no children, etc.

“Those things can really inform messaging,” Boucher says. “The further down the road of big data, the more personalized you can get with every level of your fundraising.”

It’s a team effort
“You should be aware from the outset that this requires a huge commitment on the part of the entire organization. It’s not something that one person can give it a go,” Boucher says. “And it’s not going to happen overnight.”

Boucher says it’s important to start small and test a manageable project to demonstrate results and gain momentum and support from there.

Use many methods
The organizations that are most successful with big data, Boucher says, use a lot of different methods for data collection — internal data, appended data through screening, modeled data, surveys, focus groups, donor behaviors, etc.

“Using as many different methods as possible really helps create a data environment to extract meaningful results. If you’re always relying on the same data, bias begins to creep in. It’s good to have a fresh influx of data from different sources,” she adds.